Organic Letters
Letter
observed according to the 31P NMR spectrum of the product
(Figure S26), with the major component being the enamine
(−32 ppm). Subsequent reduction gives the corresponding
NPN type 2,2′-diaminophosphine products 9a−c in approx-
imately 60% yield over the two steps, slightly higher than an
analog with a cyclopentene backbone.49 Similar to the trend
observed in phosphinoalkyne hydroamination, aromatic
amines react faster than alkyl amines in the hydroamination
step (4 h vs 18 h).
Scheme 2. Sequential Hydroamination−Reduction
a
Synthesis of 2-Aminophosphines
These results demonstrate the flexible synthesis of diverse
P,N-saturated proligands using regioselective catalytic hydro-
amination and subsequent reduction. These proligands can
then be isolated and stored for later use. These results show
that P,N-chelating ligands with strong σ- and π-donating amido
substituents, coupled with π-accepting phosphines, can be
readily accessed using this approach. Alternatively, the
intermediate phosphinoenamine/imine products of hydro-
amination can be used as ligands directly, and to date, this is
a class of chelating P,N-ligands that is largely unexplored.
Additionally, NPN type molecules could also be prepared using
the same strategy, providing an efficient synthetic route to such
kind of multidentate ligands.
One-Pot Synthesis of P,N-Ligated Catalysts. The reaction
of phosphinoalkynes in anti-Markovnikov hydroamination
generates phosphinoenamine/imine ligands in an atom-
economic approach. Most importantly, this intermediate
product that cannot be easily isolated can be used directly in
a ligand substitution reaction onto reactive metal centers. This
in situ ligand preparation avoids the stoichiometric reduction
step and offers direct access to phosphinoenamine ligands.
Related ortho-P,N substituted arenes have been used to
advantage in a range of catalytic transformations.46,49,66−69
Here we can access rarely investigated analogues without the
aromatic backbone.45−51 The handful of previous reports of
such chelating ligands reveal tedious syntheses, unlike the one-
pot method developed here.
a
Reaction time and isolated yields are listed.
Ammonia borane (H3BNH3) has been extensively studied as
a promising material for hydrogen storage due to its high
hydrogen capacity (19.6 wt %).70,71 A variety of catalysts have
been developed to facilitate dehydrogenation of ammonia
borane, which includes Fe complexes as inexpensive and earth-
abundant-metal catalysts for this desirable transformation.
Specifically, a saturated bis(2-phosphinoamido) ligated Fe(II)
complex has been developed for the dehydrogenation of
ammonia borane.72 Catalyst development efforts found that it
showed good reactivity for ammonia borane dehydropolyme-
rization; however, the ligand synthesis involved a multistep
sequence, including protection and deprotection protocols,
and could not be easily modified to accommodate diverse
substituents.72 Furthermore, low turnover numbers were
observed, due to proposed proton abstraction resulting in
catalyst decomposition. Here we show that by using a one-pot
hydroamination−deprotonation−complexation sequence com-
plementary unsaturated phosphinoenamido ligands can be
used to generate new catalysts for ammonia borane
dehydrogenation (Figure 2). Furthermore, this class of
unsaturated phosphinoenamido ligand may not be resistant
to complex degradation by deprotonation.
an almost quantitative yield of the crude product. This column-
free isolation protocol has proven to be broadly useful, and 3c
(95% yield at 1 g scale) is provided as another example.14,15
The known alkyl substituted phosphine starting material t-
Bu2P−CCH60 was less reactive requiring a longer reaction
time (48 h) and elevated temperature (120 °C) for the
reaction with 4-methylaniline. Subsequent reduction and
purification gave 4 in 72% yield. This can be attributed to
the bulky t-Bu groups on the phosphorus atom and the
reduced electronic polarization of the alkyne with the change
to an alkyl phosphine.65 However, the less reactive sec-BuNH2
showed no hydroamination reactivity, even with heating at 120
°C for 3 days.
Internal alkynylphosphines, such as Ph2P−CCPh, are
challenging substrates, demanding a one-week reaction to
S19). Further NaBH4 reduction also required gentle heating to
give aminophosphine 5 in 67% yield. Once again, only one
regioisomer was formed, which highlights the regioselectivity
of the [Ti] precatalyst in internal alkyne hydroamination.56
Less reactive alkyl-substituted internal alkynylphosphines, such
as Ph2P−CCn‑Bu did not react with 4-methylaniline under
similar conditions.
Complex 6 can be easily prepared by direct deprotonation of
the hydroamination product, intermediate phosphinoenamine
2e, using a strong base like KHMDS to give 2-phosphinoe-
namide monoanionic ligand salts. These ligand salts can then
be directly installed onto transition metals using a salt
Interestingly, by reacting 2 equiv of amine with a
dialkynylphosphine PhP(CCH)2, symmetric β,β′-phos-
phinodienamines were successfully prepared. Noteworthy, a
mixture of tautomerizing phosphino-enamine and imines was
1976
Org. Lett. 2021, 23, 1974−1979